Cognitive decline is a common component of aging, and spatial memory in particular is negatively affected by increasing age. One possible cause of impaired spatial memory is an age-related decline in circulating testosterone. Testosterone levels among men decline steadily with increasing age, and as many as 65% of men over age 70 are hypogonadal. Past studies have produced conflicting results regarding the cognitive benefits of androgen replacement therapies given to hypogonadal aged men, indicating a need for an animal model to experimentally test the effects of testosterone on spatial memory. The proposed experiments will systematically test the physiological mechanisms underlying testosterone-induced changes in spatial learning and memory, using castrated aged male rats as a model for hypogonadal aged men. The specific aims of the proposed experiments are to determine: 1) whether testosterone has dose-dependent effects on spatial memory in hypogonadal aged males, 2) whether the effects of testosterone on spatial memory are estrogen or androgen dependent, 3) whether testosterone influences spatial memory through direct actions on the brain, and 4) whether changes in brain-derived neurotrophic factor (BDNF) are involved in testosterone-induced changes in spatial memory. Three major experiments involving adult male rats will be performed to achieve these aims. Each experiment will involve hormone injections given to castrated aged male rats. For Experiments 1 and 2, spatial memory will be assessed using a working-reference memory version of the 8-arm radial maze. In Experiment 1a, aged males will be compared to young male rats to determine whether castration has the same memory impairing effects in both age groups. In Experiment 1b, a broad range of testosterone doses will be injected into aged males to assess possible dose-dependent effects. Experiment 2 will test the relative roles of androgen-dependent and estrogen-dependent pathways on spatial memory by injecting rats with various dihydrotestosterone or estradiol doses. These pathways will be further assessed using an androgen antagonist and an aromatase inhibitor. In Experiment 3, the direct effects of testosterone on spatial memory will be tested using intracranial infusions of high and low testosterone doses. Infusions into the hippocampus will be followed by memory testing on a hippocampus-dependent task, and infusions into the striatum will be followed by memory testing on a striatum-dependent task. For all experiments, serum hormone levels will be assayed, and BDNF levels will be measured in key brain regions. In combination, these experiments will provide a critical step in determining the therapeutic value of androgen replacement therapies for treating age-associated memory impairment.